Fly ash based geopolymeric foams using silica fume as pore generation agent. Physical, mechanical and acoustic properties

Fly ash based geopolymeric foams using silica fume as pore generation agent. Physical, 1 mechanical and acoustic properties 2 *Y. Luna-Galiano, C. Leiva, C. Arenas, C. Fernández-Pereira 3 University of Seville, School of Engineering, Chemical and Environmental Engineering 4 Department, Camino de los Descubrimientos s/n 41092, Seville (Spain), (0034-954481180, Fax: 5 0034-954486082, [email protected]) 6 Abstract 7 The aim of this work is the development of a porous geopolymeric foam with sound absorbing 8 properties using silica fume as the pore generation agent. The samples were manufactured using a 9 coal combustion fly ash as source material for the geopolymerization reaction, an alkaline solution 10 as activating solution and silica fume as pore forming agent. Three parameters were studied: silica 11 fume proportion (0, 20 and 40 wt%), activating solution (potassium silicate and potassium 12 hydroxide) and setting temperature (40 and 70 ºC). Once the samples were prepared, the 13 experimental study of the most important physical, mechanical and acoustic features were carried 14 out. The increase in the proportion of silica fume in the mixture and setting temperature produced a 15 reduction in setting time and a raise of open void porosity, reducing the compressive strength at 28 16 days up to a half but increasing the sound absorption. The influence of activating solution in open 17 porosity and sound absorbing properties was not very important. 18 19 Keywords: fly ash, foam geopolymer, pore generation, open porosity, mechanical properties, 20 acoustic properties 21 1. Introduction 22 Licencia Creative Commons CC BY-NC-ND Geopolymers are considered an eco-friendly construction material [1]. This material exhibits great 23 potential in several applications [2]: as binder agent in the stabilization/solidification of industrial 24 wastes [3-6], as fire resistant material in construction [7-10], as hydraulic road binder [11], as 25 acoustic concrete [12-13], among others [14]. The construction field is where geopolymers can 26 compete with cement products [15] due to a large number of useful/beneficial properties: high 27 compressive strength and structural integrity [16], resistance to acids and salts [16, 17], low 28 permeability [16], and medium-high density [18]. The challenge of the current work is to study and 29 evaluate pore generation in a fly ash based geopolymer in order to manufacture a foam product. A 30 foam product is considered a type of lightweight material that consists of a binder with a high 31 degree of void space [19]. There are two traditional methods of pore generation in a mortar [19]: by 32 endogenous gas generation, which can be achieved by mixing gas-releasing agents such as H2O2 33 [20], aluminium [21, 22] or silica [23, 24], or by introducing a very large volume fraction of air 34 bubbles, normally using an organic foaming agent (surfactants) [20, 25, 26]. Silicon, in the form of 35 silica fume, was used to generate pores in this work. In general, a reactive metal powder (Si) reacts 36 with water and hydroxide in an alkaline environment producing bubbles of hydrogen gas [19]. 37 These bubbles attempt to escape into the air from the geopolymer paste producing the expansion of 38 the geopolymer mortar [27] and generating the porous structure. Silica fume is defined as an 39 inorganic by-product from the metallurgical industry [28]. Several studies have been carried out in 40 which silica fume is added to a metakaolin based geopolymer [24, 29, 30] in order to generate a 41 foam. Prud´homme et al., [23, 24] studied the effect of silica fume content and temperature in the 42 microstructure and porosity of geopolymer foam. Henon [30] concluded that the synthesis of the 43 product and porosity depend to a great extent on the temperature (range 25-70 ºC). In his study, 44 geopolymer foams with macroporosities in the range 05-1.6 mm were developed. Thermal 45 conductivities of 0.12-0.33 W/(mK) for porosities of 65-85 % were obtained. Papa et al., [29] 46 obtained macroporous structures with roughly rounded pores and a total porosity of about 80%, 47 thermal conductivities of around 0.16 W/(mK) and acoustic behaviour. 48 In the current work, fly ash based geopolymers have been prepared adding silica fume in different 49 proportions in order to evaluate the physical, mechanical and acoustical properties of the final 50 geopolymeric foam. Two activating solutions and setting temperatures have also been evaluated. In 51 summary, a dual environmental design has been reached: to recycle industrial by-products produced 52 in large amounts and to develop an acoustic absorbing material in order to reduce noise pollution 53 levels. 54 55 2. Materials and methods 56 2.1 Materials 57 A low calcium fly ash (FA) (ASTM class F [31]) from a coal combustion power plant (Los Barrios, 58 Spain) was used as the geopolymeric precursor. The silica fume (SF) used in this work was supplied 59 by the Sika© company. Chemical composition of the fly ash and the silica fume (Table 1) was 60 determined after chemical attack and dissolution at 750 ºC (ASTM D-3682-78) [32] using atomic 61 absorption spectroscopy. The specific gravity, in accordance with EN 1097-7 [33], was also 62 determined in the fly ash and silica fume (Table 1). 63 Table 1. Characteristics of fly ash and silica fume 64 Moisture (%) LOI 750ºC Chemical composition (wt %) Specific gravity (g/cm3) Fe2O3 CaO MgO SiO2 Al2O3 Na2O K2O FA 0,05 3,32 5,86 3,94 1,84 63,9 21,5 0,68 1,67 1,93 SF 0.03 1,48 0,5 0,51 0,39 96,7 0,43 0,2 0,79 2,04 65 High SiO2 and Al2O3 contents were observed in FA. The silica fume was practically SiO2, and its 66 specific gravity (2.04 g/cm3) was similar to the fly ash (1.93 g/cm3). Amorphous contents of both 67 materials were determined after the X-ray powder diffraction (XRD) analysis using DIFFRAC.EVA 68 software. Amorphous content of FA and SF were 44.1 % and 67.1 %, respectively. 69 The particle size distribution of both materials was determined using a laser diffraction 70 granulometer provided by Micromeritics, which is equipped with an optical system and data 71 acquisition software. The results are shown in Figure 1. 72 Figure 1. Particle size distribution of fly ash and silica fume 73 74 As can be seen in Figure 1, FA presented the finest particle size distribution, with 88 % of particles 75 less than 45 µm. SF showed a coarser distribution than FA, with 80 % of particles lower than 100 76 µm. 77 Two activating solutions were used in this work. The first solution was a mixture of potassium 78 silicate solution (SiO2 23 w/w % and K2O 14.9 w/w %) (Industrias Químicas del Ebro, Spain) and 79 sufficient potassium hydroxide to increase the K2O/SiO2 molar ratio up to 1.45 (pH = 13.5). The 80 second solution was KOH 8M (pH = 14). 81 2.2 Geopolymeric foam preparation 82 Geopolymeric foams were manufactured in a laboratory mixer working at 500 rpm. Fly ash was 83 mixed with an activating solution until a thixotropic paste was obtained (mixing time was 4 84 minutes). After that, silica fume was added to the geopolymer paste. A reduction in thixotropy and 85 workability of the geopolymers was observed as SF was mixed with the geopolymer paste, so extra 86 water was necessary to improve both characteristics. The pastes obtained were poured into 87 cylindrical plastic moulds (30 mm-diameter and 40 mm-height). Two setting temperatures were 88 studied: 40 ºC and 70 ºC (in the first 24 hours). After setting, samples were demoulded and cured at 89 room temperature (average temperature: 20 ºC; average relative humidity: 45 %) for a further 27 90 days. Compositions of geopolymeric foams are detailed in Table 2. The samples were cut after 28 91 days in order to obtain the same sample thickness (40 mm) in the different tests. 92 Table 2. Composition of geopolymeric foam 93 Geopolymer foam Solid phase (wt%) Activating solution (g)/Fly ash (g) Water (g)/Solid phase (g) Setting temperature (ºC) FA SF FA100SF0-KSil-40ºC 100 0 0,74 0 40 FA80SF20-KSil-40ºC 80 20 0,74 0,1 40 FA60SF40-KSil-40ºC 60 40 0,74 0,225 40 FA80SF20-KSil-70ºC 80 20 0,74 10 70 FA80SF20-KOH-40ºC 80 20 0,71 0 40 94 As can be seen in Table 2, three parameters have been studied: FA-SF proportion (100-0, 80-20, 60- 95 40), activating solution (potassium silicate (KSil) vs potassium hydroxide (KOH)), and setting 96 temperature (40 vs 70 ºC). As the silica fume content increased in the mixture, the water/solid ratio 97 also increased. This phenomenon suggests that SF may demonstrate inert material behaviour in the 98 geopolymerization reaction. As the addition of SF produced a foam material, final products 99 obtained from samples FA80SF20 and FA60SF40 were also named geopolymeric foams. However, 100 the final product of the sample FA100SF0 was a geopolymeric paste. 101 2.3. Methods 102 2.3.1. Volume expansion and setting time 103 The geopolymerization reaction kinetic was studied analysing the setting time. The setting time was 104 determined using a Vicat needle (EN 196-3) [34]. The initial setting time is the time elapsed 105 between the initial instant and the time in which the distance from the needle to the bottom base of 106 the cylinder is 6±3 mm. The final setting time is the time between the initial instant and the time in 107 which the needle shows no visible signal on the upper base of the cylinder. The pore generation 108 reaction kinetic was determined by measuring the expansion volume (EV) of material during setting 109 (mm). EV is the ratio between the foam volume at instant t, V(t), and the initial volume of 110 introduced paste, V(0). For this test, geopolymer samples were moulded in 35 mm-diameter and 80 111 mm-high cyclindrical plastic molds no more than half-full. Three specimens of each geopolymer 112 mixtures were used to determine the volume expansion and setting time. 113 2.3.2. XRD analysis 114 Geopolymer samples were studied by means of X-ray powder diffraction (XRD) to evaluate their 115 mineralogical composition. A D8 Advance A25 (BRUKER) (40 kV and 30 mA) instrument was 116 used and phase identification was carried out using DIFFRAC.EVA software (BRUKER). 117 2.3.3. Physical and mechanical properties 118 Density () of the geopolymers was measured in accordance with the EN 1936 standard method 119 [35]. Open void ratio (open porosity) was also determined because it is strongly related to the 120 acoustic behaviour of the products. The method of vacuum water saturation was followed in the 121 determination of open porosity (OP %) [35]. 122 Compressive strength of geopolymer samples was determined in accordance with EN 196-1 [36], 123 using a Tinius Olsen-TO317EDG machine. Four specimens of each geopolymer (paste or foam) 124 were used to determine the compressive strength, density and open porosity. 125 2.3.4. Acoustic properties 126 In order to evaluate the acoustic properties of the geopolymeric samples, sound absorption was 127 determined using a Kundt tube, according to EN ISO 10534-2 [37]. An ACUPRO system, 128 implemented by “Spectronics Inc.”, with two microphones and a SAMSON signal amplifier, was 129 used. Three specimens of each geopolymer have been tested. The thickness of all the tested 130 specimens was 40 mm. The Noise Reduction Coefficient (NRC) was determined by calculating the 131 arithmetic mean of the absorption coefficients at 250 Hz, 500 Hz, 1000 Hz and 2000 Hz. 132 133 3. RESULTS 134 3.1. Expansion volume and setting evolution with time 135 Figures 2, 3 and 4 show the setting curves (penetration of Vicat needle (mm) in the cylinder versus 136 time) (Set) and the expansion volume curves (expansion volume data (natural logarithm) (EV) 137 versus time), taking into account the three parameters studied: FA-SF proportion (100-0, 80-20, 60- 138 40) in Figure 2, activating solution (potassium silicate vs potassium hydroxide) in Figure 3 and 139 setting temperature (40 vs 70 ºC) in Figure 4. Standard deviations are also represented as error bars. 140 Figure 2. Expansion volume and setting curves for different FA-SF proportions 141 142 143 144 Figure 2 shows that the substitution of FA with SF produced a reduction in the initial setting time 145 (from 1410 min for geopolymer paste without silica fume (FA100SF0) to 120 min for geopolymer 146 foam with 40 %wt of silica fume (FA60SF40)) and final setting time (from 1620 min without silica 147 fume to 450 min with 40 % wt of silica fume). Figure 2 also shows an increment in material 148 expansion volume as silica fume was added to the mixture due to generation of gas produced by the 149 use of SF (as SF content was null, EV was null). 150 Figure 3. Volume expansion and setting curves for different activating solutions (potassium 151 silicate vs potassium hydroxide) 152 153 As can be seen in Figure 3, the change of activating solution did not bring about important 154 variations in setting time and material expansion volume results. A slight reduction in the setting 155 time in geopolymer foam activated with potassium silicate in comparison with the activation with 156 potassium hydroxide was observed. A slight increment in expansion volume as geopolymer foam 157 was activated with potassium hydroxide can also be observed. 158 Figure 4. Volume expansion and setting curves for different setting temperatures (40 vs 70 ºC) 159 As can 227 be seen in Figure 10, both curves were extremely similar (the curve of geopolymer foam activated 228 with hydroxide was marginally above the curve of geopolymer foam activated with silicate). Both 229 geopolymer foams had a similar open porosity (slightly higher open porosity was observed in the 230 geopolymer foam activated with hydroxide), which may explain the similarity between the curves. 231 Figure 11. Sound absorption coefficient. Setting temperature (40 vs 70 ºC) 232 233 Figure 11 shows that both curves had a similar shape. Two peaks at 400 and 2500 Hz were also 234 observed. The highest values of the sound absorption coefficient were observed in the geopolymer 235 foam set at 70 ºC since this material presented the greatest open porosity. 236 Noise Reduction Coefficient (NRC) for all geopolymers is detailed in Table 4. This coefficient is a 237 numerical descriptor of the sound absorption, which is very useful to make comparisons. Standard 238 deviations are also detailed. Analysing Table 4 and Table 3 we can see that the NRC increased as 239 porosity increased. 240 Table 4. NRC of geopolymer foams 241 Geopolymer foam NRC FA100SF0-KSil-40ºC 0.08±0.01 FA80SF20-KSil-40ºC 0.14±0.01 FA60SF40-KSil-40ºC 0.21±0.02 FA80SF20-KSil-70ºC 0.23±0.02 FA80SF20-KOH-40ºC 0.16±0.01 242 4. DISCUSSION 243 4.1. Volume expansion and setting time 244 Two reactions take place during the geopolymer foam manufacture: the geopolymerization reaction 245 and the pore generation reaction. The geopolymerization reaction occurs when the fly ash reacts 246 with the activating solution, producing an alkali aluminosilicate hydrated gel. The gel obtained in 247 the current work was mainly based on potassium (potassium aluminate silicate hydrated gel) since 248 the activating solution contains potassium as alkali cation [38]. The use of different types of anions 249 (silicate or hydroxides) also affects the final chemical composition of the gel (Si/Al ratios are higher 250 in gels after activation with silicate than with hydroxide) [38, 39]. The pore generation is due to the 251 H2(g) released during the water reduction and silicon (from silica fume) oxidation, producing 252 Si(OH)4. The reaction is promoted in basic medium [24] according to: 253 4H2O + Si →2H2 (g) + Si(OH)4 (eq 1) 254 As can be seen in Figures 2, 3 and 4, setting curves exhibited three clearly differentiated stages. The 255 first stage corresponded to the plastic period of the geopolymer, which goes from the time in which 256 needle penetrates completely to the start of setting. The second stage matched with the setting 257 period of the geopolymer, which is between initial and final setting time (needle shows resistance to 258 penetration). The final stage constituted the hardening period. 259 Expansion volume curves also showed three stages. The first stage began with a period of latency 260 during which the expansion volume was null, named latency time [30]. The second stage of the 261 curve exhibited an increase of expansion volume (EV) with the time caused by the generation of 262 gas, which follows a linear regime [30]. The last showed a final expansion volume constant through 263 time. 264 The replacement of fly ash with silica fume produced a reduction in the initial and final setting time 265 of the geopolymer, that is, the geopolymer paste FA100SF0 set more slowly than the geopolymer 266 foam FA80SF20 and the geopolymer foam FA80SF20 more slowly than FA60SF40. This behavior 267 is similar in silica fume blended cement pastes [40] or in silica fume blended metakaolin 268 geopolymers [24] where an increment in silica fume accelerated the setting and hardening process. 269 An increment in material expansion volume as silica fume was added to the mixture was also 270 observed. This is due to the generation of gas produced by the use of SF (according to equation 1). 271 In addition, the latency period matched the plastic period and the setting time coincided with the 272 period in which gas was generating. As gas is generated, the hydrogen bubbles try to escape from 273 the material. When the geopolymer is still fluid in the plastic period (setting has not yet taken 274 place), the gas bubbles generated can leave the geopolymer. However, while the geopolymer is 275 setting and the material is hardening, the gas becomes trapped in the structure, expanding the 276 geopolymeric matrix and increasing the expansion volume. 277 The type of activating solution did not significantly affect the setting time or the expansion volume. 278 A slight reduction in the setting time of geopolymer foam activated with potassium silicate 279 compared with that activated with potassium hydroxide was observed. Some authors [39] noted that 280 geopolymers activated with soluble silicate showed higher reaction rates than those activated with 281 hydroxide. Furthermore, previous work of the authors [41], where the activation with hydroxide 282 versus silicate was studied, concluded that systems activated with hydroxide (KOH 8M) presented 283 (fewer) lower degrees of reaction, at the same time, than systems activated with silicates (potassium 284 silicate), which could provide evidence of the high rate of reaction with silicates. Furthermore, pH 285 of the activating solution must be also considered. Some authors [42] revealed that high 286 concentration solutions presented a coagulation of silica, reducing the dissolution of Si4+ and Al3+ 287 from the raw material. In the current work, pH of potassium hydroxide solution was >14 and pH of 288 potassium silicate solution was 13.5. Therefore, the increase in setting time using KOH 8M could 289 be the result of some silica coagulation, which may result in some setting delay. 290 Material expansion volume was slightly higher in geopolymer activated with potassium hydroxide, 291 possibly due to the different solution pH (pH>14 in the potassium hydroxide solution and pH = 13.5 292 in the potassium silicate solution). As previously mentioned, the pore generation reaction occurs in 293 basic medium, so the addition of a more basic solution could have a positive effect on this reaction. 294 The effect of temperature is also displayed, showing a reduction in setting time and an increment in 295 material expansion as geopolymer foam samples set at 70 ºC instead of 40 ºC. On one hand, it is 296 well known that temperature is a geopolymerization reaction accelerator [39]. On the other hand, 297 the reaction of equation 1 is enhanced by the temperature [43]. Henon et al. [30] have proposed that 298 the final expansion of metakaolin based geopolymer with silica fume followed an Arrhenius law 299 with the temperature, obtaining a value of activation energy of 11.9 kJ/mol. The activating energy 300 of geopolymer foam FA80SF20-KSil in the current work was also calculated according to the 301 methodology of Henon’s work and the value was 27.9 kJ/mol. This value was similar to that 302 obtained in a study carried out by Provis [44] (33 kJ/mol) where metakaolin was used as 303 geopolymeric raw material. 304 4.2. XRD analysis 305 XRD results show the changes which took place in the SF and the FA after the geopolymerization 306 reaction. As previously mentioned, FA displayed a slight broad reflection in the 2θ range 15-38 307 with crystalline peaks of quartz and mullite. The geopolymer paste FA100SF0-KSil showed a shift 308 of the broad peak in the 2θ range 22-38 regarding the FA pattern, which provides evidence of the 309 dissolution of SiO4 and AlO4 species during the geopolymerization reaction [24]. In addition, the 310 crystalline peaks in the geopolymer (quartz and mullite) were in the same 2θ position as the FA 311 pattern but with lower intensity, which was indicative of alteration of the crystal structure [23]. SF 312 pattern was characteristic of an amorphous material, with a broad peak, which was not found in the 313 geopolymer foam FA60SF40-KSil pattern. The geopolymer foam pattern was rather similar to the 314 geopolymer paste and the FA patterns, with a slight shift of the broad peak in the 2θ position and 315 with the crystalline peaks (quartz and mullite) in the same place but with lower intensity. 316 Comparing the XRD patterns of the geopolymer paste and the geopolymer foam we can see that the 317 silica fume addition did not produce significant differences in amorphous and crystalline phases, 318 except a slight difference in the intensity of the peaks. Amorphous content of both materials were 319 calculated using DIFRACT.EVA software to support these results. FA100SF0 and FA60SF40 320 showed values of 51.6 % and 53.8 %, respectively. 321 4.3. Physical and mechanical properties 322 Table 3 and Figures 2, 3 and 4 provide evidence of the direct relationship between open porosity 323 and expansion volume of the material. The material expansion test (Figure 2) shows higher foam 324 development of the geopolymer as the proportion of silica fume increased, which allows us to view 325 the positive effect of silica fume proportion in pore generation, and consequently, in the open 326 porosity. The same behavior was displayed as the activating solution effect was analyzed. 327 Geopolymer foam activated with hydroxide (Figure 3) showed a slightly higher expansion volume 328 than that activated with potassium silicate, so the amount of generated pore was slightly greater, as 329 was the open porosity. The values of open porosity depend greatly on the setting temperature, 330 producing a high increment in open porosity as the temperature increased from 40 to 70 ºC due to 331 the increment in the expansion volume (Figure 4). Photographs of different geopolymer foams are 332 detailed in Figure 12. Analyzing the pore size in the right photograph, it can be observed that pore 333 sizes were in the range 1-3 mm, which are the most appropriate pore sizes for noise absorbing 334 materials [45]. 335 Figure 12. View of the some of the geopolymer foam (A. FA100SF0-40ºC; B. FA60SF40-40ºC; 336 C. FA80SF20-40ºC; D. FA80SF20-70ºC) after 28 days of curing 337 338 The effect of silica fume content, activating solution and setting temperature on compressive 339 strength depend on the curing time in which the test was carried out: early stages (1 day) or final 340 stages (28 days). It seems that compressive strength at early stages was controlled by the setting 341 rate. Therefore, variables that produce quick setting times such as using a high content of silica 342 fume (up to 40%), to set at 70 ºC and to activate with potassium silicate, can improve the 343 compressive strength at 1 day of curing. However, the compressive strength at final stages was 344 controlled by the open porosity values. The parameters that reduced the open porosity (no use of 345 silica fume, to set at 40 ºC and to activate with potassium silicate) also produced an enhancement of 346 the compressive strength at 28 days of curing. It is well known that an increase in voids or pores in 347 any material (while the rest of the parameters remain unchanged) always causes a reduction in the 348 compressive strength since the volume/resistant mass ratio is lower [46], so the poorest mechanical 349 properties were observed in the most porous geopolymer foams. 350 4.4. Acoustic properties 351 Taking into account that sound absorption in a porous material is related to the energy loss through 352 friction with the wall of the holes, higher porous structures showed higher sound absorbing 353 properties [12]. Therefore, to increase the silica fume proportion and to increase the setting 354 temperature from 40 to 70 ºC produced a positive effect on sound absorption. In addition, these 355 geopolymer foams exhibited sound absorption coefficients similar to or better than commercial 356 porous concrete [46] and other sound absorption materials containing wastes, such as bottom ashes 357 [46] or ceramic wastes [47]. 358 359 5. CONCLUSIONS 360 The main aim of this experimental study was to develop geopolymeric foams using fly ash and 361 silica fume. The foam products presented porous material characteristics with relatively good noise 362 absorption properties. Three parameters have been studied: silica fume content, activating solution 363 and setting temperature. Five properties have been evaluated: setting time, expansion volume, open 364 porosity, compressive strength (at early and final stages) and sound absorption, all of which are 365 closely linked. 366 The substitution of fly ash with silica fume and the set at 70 ºC reduced the initial and final setting 367 time and increased material expansion. The first observed effect was an increment in the open 368 porosity of the geopolymer foam. Taking into account that the sound absorption in a porous 369 material was related to the energy loss through friction with the walls of the holes, less porous 370 structures showed lower sound absorbing properties. Therefore, geopolymer foam prepared with the 371 highest amount of silica fume and set at 70 ºC showed the greatest noise reduction coefficient. The 372 effect of the three parameters on compressive strength depends on the curing time. The behaviour of 373 the materials at early and final stages of curing is not the same since the compressive strength 374 values at early stages (1 day) were controlled by the setting rate. However, the compressive strength 375 values at 28 days were directly related to the open porosity values. 376 It seems that to replace one activating solution for another did not produce significant differences. A 377 slight reduction in setting time using silicate, and a slight increment in open porosity (expansion 378 volume) using hydroxide was observed, possibly due to the variations in the pH values of solutions 379 and the different degrees of reaction reached by the different material products obtained. 380 381 Compliance with Ethical Standards: 382 Authors declare that they have no conflicts of interest. 383 384 Acknowledgements: 385 Authors would like acknowledge the help received by CITIUS (General Research Services) from 386 University of Seville, especially the XRD laboratory. 387 388 References 389 [1] Zhang L, Zhang F, Liu M, Hu X. Novel sustainable geopolymer based syntactic foams: An eco- 390 friendly alternative to polymer based syntactic foams. Chem. Eng. J. 313 (2017) 74-82. DOI: 391 10.1016/j.cej.2016.12.046. 392 [2] Palomo A, Krivenko P, Garcia-Lodeiro I, Kavalerova E, Maltseva O, Fernández-Jiménez A. A 393 review on alkaline activation: new analytical perspectives. Mater. Construcc. 64(315) (2014) e022. 394 https://doi.org/10.3989/mc.2014.00314. 395 [3] Yunsheng Z, Wei S, Qianli C, Lin C. Synthesis and heavy metal immobilization behaviours of 396 slag based geopolymer. J. Hazard. Mater. 143(1–2) (2007) 206–213. DOI: 10.1016/j.jhazmat.2006. 397 09.033 398 [4] Luna-Galiano Y, Fernandez Pereira C, Vale J. Stabilization/solidification of a municipal solid 399 waste incineration residue using fly ash-based geopolymers. J. Hazard. Mater. 185(1) (2011) 373– 400 381. DOI: 10.1016/j.jhazmat.2010.08.127 401 [5] Zhang J, Provis JL, Feng D, van Deventer JSJ. Geopolymers for immobilization of Cr(6+), 402 Cd(2+), and Pb(2+). J. Hazard. Mater. 157(2–3) (2008) 587–598. DOI: 10.1016/j.jhazmat.2008. 01. 403 053 404 [6] Fernández Pereira C, Luna-Galiano Y, Querol X, Antenucci D., Vale J. Waste 405 stabilization/solidification of an electric arc furnace dust using fly ash-based geopolymers. Fuel. 406 88(7) (2009) 1185-1193. DOI: 10.1016/j.fuel.2008.01.021. 407 [7] Zhang HY, Kodur V, Qi SL, Cao L, Wu B. Development of metakaolin–fly ash based 408 geopolymers for fire resistance applications. Constr. Build. Mater. 55 (2014) 38-45. DOI: 409 10.1016/j.conbuildmat.2014.01.040 410 [8] Zhang HY, Kodur V, Qi SL, Wu B, Cao L, Fan Wang. Thermal behavior and mechanical 411 properties of geopolymer mortar after exposure to elevated temperatures. Constr. Build. Mater. 109 412 (2016) 17-24. DOI: 10.1016/j.conbuildmat.2016.01.043 413 [9] Luna-Galiano Y, Cornejo A, Leiva C, Vilches LF, Fernández-Pereira C. Properties of fly ash 414 and metakaolín based geopolymer panels under fire resistance tests. Mater. Construcc. 65(319) 415 (2015) e059. http://dx.doi.org/10.3989/mc.2015.06114. 416 [10] Luna-Galiano Y, Leiva C, Arenas C, Arroyo F, Vilches LF, Fernández Pereira C, Villegas R. 417 Behaviour of Fly Ash-Based Geopolymer Panels Under Fire. Waste. Biomass. Valor. 8(7) (2017) 418 2485-2494. DOI: 10.1007/s12649-016-9803-y 419 [11] Luna-Galiano Y, Arenas C, Cornejo A, Leiva C, Vilches LF, Fernández-Pereira C. Recycling 420 by-products from coal-fired power stations into different construction materials. Int. J. Energ. 421 Environ. Eng. 5 (2014) 387–397. DOI: 10.1007/s40095-014-0120-6 422 [12] Arenas C, Luna-Galiano Y, Leiva C, Vilches LF, Arroyo F, Villegas R, Fernández-Pereira C. 423 Development of a fly ash-based geopolymeric concrete with construction and demolition wastes as 424 aggregates in acoustic barriers. Constr. Build. Mater. 134 (2017) 433–442. https://doi.org/10.1016 425 /j.conbuildmat.2016.12.119 426